Tariffs are described in political argument in ways that conflict with the economics, and the mechanism is straightforward enough to set out.
The mechanics
A tariff is a tax on imported goods, collected at the border from the importer.
Which means the legal incidence falls on the domestic importing business, not on the exporting country.
The economic incidence — who ultimately bears the cost — depends on how prices adjust, which is the empirical question.
Pass-through
Studies of recent tariff episodes have generally found high pass-through to domestic prices.
Which means domestic buyers paid most of the cost through higher prices, rather than foreign exporters absorbing it through lower prices.
This has been found repeatedly in careful empirical work on several episodes, and it is one of the more consistent findings in trade economics.
Pass-through is lower where the importing country is a large enough share of the exporter's market to force price concessions.
Effects on domestic producers
Domestic producers competing with the imported goods benefit from reduced competition.
Domestic producers using the imported goods as inputs face higher costs.
Which means tariffs on intermediate goods harm downstream manufacturers, and this effect frequently exceeds the benefit to the protected sector in employment terms.
Studies of specific tariff programmes have found employment gains in protected sectors offset by larger losses in user industries.
Retaliation
Trading partners generally respond with tariffs on exports from the imposing country.
Which is targeted strategically, frequently at politically sensitive sectors.
Agricultural exports have been a common target, and support payments to affected farmers have followed in several cases, which transfers the cost to taxpayers.
Exchange rates
Tariffs can affect currency values, which offsets part of the price effect.
Which is a theoretical channel with mixed empirical support, and the offset in observed episodes has generally been partial.
The revenue question
Tariffs raise revenue, and the amounts relative to total government revenue are small in most developed economies.
Which was not always true — tariffs were a primary revenue source historically, before income taxation.
Very high tariffs raise less than proportionally, since they reduce the imports being taxed.
Non-tariff barriers
Standards, licensing, quotas and procedural requirements can restrict trade as effectively.
Which are harder to measure and are the main subject of modern trade negotiation, since tariffs are already low between developed economies.
The political economy
Benefits of protection are concentrated in a visible sector. Costs are dispersed across all consumers and across user industries.
Which produces asymmetric political pressure, and it explains why protection is politically easier to impose than to remove.
This is the classic collective action problem applied to trade, and it has been described in the literature for decades.
Where economists disagree
On strategic industries, national security exceptions, responding to unfair practices by trading partners, and adjustment support for affected workers.
Which are genuine debates, and the pass-through and incidence findings are not among them.
Rules of origin
Determining where a product originates for tariff purposes, which is complicated by supply chains crossing many borders.
Which requires rules specifying how much value must be added locally, or which processing steps determine origin.
Compliance requires documentation through the chain, and the administrative burden is substantial enough that some firms pay the tariff rather than claim a preference.
Exclusions and exemptions
Processes allowing specific products to be exempted from tariffs, generally on the basis that no domestic alternative exists.
Which creates a lobbying opportunity and generates substantial activity, since exclusion is worth a great deal to affected importers.
Studies of exclusion processes have found outcomes associated with political factors alongside stated criteria.
Duration
Tariffs imposed as temporary measures have frequently persisted, since removal faces the same asymmetric political pressure as imposition faced in reverse.
Which means the political economy operates in both directions, and sunset provisions requiring affirmative renewal are the structural response.
Adjustment support
Programmes assisting workers and communities affected by trade.
Which exist in several countries and have been consistently under-resourced relative to what economic models assumed would accompany liberalisation.
Evaluations of such programmes generally find modest effects, with the difficulty being that displaced workers in affected regions face limited alternative employment locally.
Regional effects have been found to persist for decades in several studies, which contradicts models assuming rapid reallocation.
Trade agreements and tariffs
Preferential agreements reduce tariffs between parties while maintaining them against others.
Which can divert trade toward less efficient partners inside the agreement, an effect identified in the theoretical literature and measured in practice.
Whether an agreement is net beneficial depends on the balance between trade creation and diversion.
Following the data
Trade statistics, tariff schedules and applied rates are published by national authorities and by international organisations.
Which allows claims about trade balances and tariff levels to be checked directly.
Bilateral trade balances in particular are frequently misinterpreted, since they reflect specialisation rather than any measure of fairness, and a country's overall balance is determined by savings and investment rather than by trade policy.